mirror of
https://github.com/wassname/pytorch-lightning.git
synced 2026-09-13 12:50:26 +08:00
Gradient accumulation callback (#150)
* Gradient accumulation callback * little test case * typo * import fix * method name fix * fix epochs indexing from 1 * better code style * code style fix v2 :/ * change interface * fix Trainre new api in tests * trainer api bug fix * new raising error, new update method * extentions tests * a little better tests * typo fix * flack8 better * using scheduler for int and dict * typo * firs epoch bug fix * test update * empty dict exception * floats check * codestyle fix * grad counting test * someday, i will install normal linter * add more checks * Update test_models.py * Update test_models.py * Update test_models.py * Update test_models.py * Update test_models.py * Update test_models.py * Update test_models.py
This commit is contained in:
committed by
William Falcon
parent
c2247350bb
commit
73cf47112e
@@ -1,6 +1,7 @@
|
||||
from .pt_callbacks import EarlyStopping, ModelCheckpoint
|
||||
from .pt_callbacks import EarlyStopping, ModelCheckpoint, GradientAccumulationScheduler
|
||||
|
||||
__all__ = [
|
||||
'EarlyStopping',
|
||||
'ModelCheckpoint',
|
||||
'GradientAccumulationScheduler',
|
||||
]
|
||||
|
||||
@@ -1,5 +1,6 @@
|
||||
import os
|
||||
import shutil
|
||||
import warnings
|
||||
|
||||
import numpy as np
|
||||
|
||||
@@ -254,6 +255,37 @@ class ModelCheckpoint(Callback):
|
||||
self.save_model(filepath, overwrite=False)
|
||||
|
||||
|
||||
class GradientAccumulationScheduler(Callback):
|
||||
"""Change gradient accumulation factor according to scheduling.
|
||||
# Arguments
|
||||
scheduling: dict, scheduling in format {epoch: accumulation_factor}
|
||||
"""
|
||||
def __init__(self, scheduling: dict):
|
||||
if scheduling == {}: # empty dict error
|
||||
raise TypeError("Empty dict cannot be interpreted correct")
|
||||
|
||||
for key in scheduling.keys():
|
||||
if not isinstance(key, int) or not isinstance(scheduling[key], int):
|
||||
raise TypeError("All epoches and accumulation factor must be integers")
|
||||
|
||||
minimal_epoch = min(scheduling.keys())
|
||||
if minimal_epoch < 1:
|
||||
msg = f"Epochs indexing from 1, epoch {minimal_epoch} cannot be interpreted correct"
|
||||
raise IndexError(msg)
|
||||
elif minimal_epoch != 1: # if user didnt define first epoch accumulation factor
|
||||
scheduling.update({1: 1})
|
||||
|
||||
self.scheduling = scheduling
|
||||
self.epochs = sorted(scheduling.keys())
|
||||
|
||||
def on_epoch_begin(self, epoch, trainer):
|
||||
epoch += 1 # indexing epochs from 1
|
||||
for i in reversed(range(len(self.epochs))):
|
||||
if epoch >= self.epochs[i]:
|
||||
trainer.accumulate_grad_batches = self.scheduling.get(self.epochs[i])
|
||||
break
|
||||
|
||||
|
||||
if __name__ == '__main__':
|
||||
c = EarlyStopping(min_delta=0.9, patience=2, verbose=True)
|
||||
losses = [10, 9, 8, 8, 6, 4.3, 5, 4.4, 2.8, 2.5]
|
||||
|
||||
Reference in New Issue
Block a user